Test 4 - Moving footage

I have returned to a moving piece of footage to see if there is any way to fix the problem I had of the light moving with the footage. I came up with this. Although not in anyway perfect or even good, it works, and is the first step in a long road of testing.

To achieve this effect, I shrunk the footage, and moved it as the camera moved, aiming to keep the background static in every shot. I then created the light and made it follow the subject. As these warped perspective, I had to try to warp it back using the warp tool, and setting two key frames, one as the camera began to move, and another at the end. I also copied this effect on the light, although this is not very obvious. To improve on this effect, I need to use motion tracking on a rendered version of the film to track the square of film wherever it goes, and make it static, and center it. This should then look better. However, I could not get the motion tracking to hold, so will hopefully try again later, or find a tutorial off the Internet.

Test 3!

This time, I have used all the past effects I have experimented with together, as well as some new ones, to actually create something. Watch it and leave a comment.



I originally had 30 odd seconds of footage to place the light over, and decided that it would be boring, so sped it up using the 'time remap' tool. I then put the light in, and chose a solid orange colour, which combined with quickly changing the brightness/contrast inspired me to slowly change the colour scheme of the film, making it look like I was watching the sun go down in fast motion. I think the finished effect is quite convincing, at least at a first glance, but due to lack of footage, the sun is clearly moving too fast, and is too low. It is also much too short - a longer version would have shown off the effects better. Either way, I am happy with the result, and this proves the original point of this, which was to check if the effect would work on footage with still camera angles. Be nice if one of you nice Internet people commented to agree/disagree though.

Tests 2

Three more videos! Go superior processing power!

This one is a short test of an improved version of the light streak, again on moving footage. I am still trying to figure out how to make the light not follow the camera, and should have a test of it next.

This is a longer test to check how well the effect works on still image/films. It looks very realistic, and although there are a could of things to iron out, the effect is very convincing.

Another, even longer test using some slightly changed settings - mainly the life of the light, which now fades slowly.

It has occured to me that there are going to be limitations to this effect given my skill with the program, and so I will need to come up with a list of these, and base the narrative/ideas around it, seeing as this gimmick is the main part of the idea.

Slow mo test.

This is a test of a slow motion effect using after effects. This was nearly entirely successful, and there are few things I would change. This was a useful technique to learn.

EDIT: I have decided to use this regularly to stretch clips to the beat, as well as to fit the mood and make tricks more interesting.

Test 1

This video is my first test for my idea. It is greatly flawed, and I hope to improve this tenfold by the time it is finished, but it is a start. It took a surprisingly small amount of time to create, and was easier than I expected it to be. However, there are still many problems with the video. The first of these is the quality of the light. It looks a bit too transparent and has no texture. It also moves too fast, and disappears too quickly and suddenly. However, all of these things can be easily changed, and I will do so in my next test video.

I also discovered some problems with the technique used when creating this. Firstly, it is much harder to make the light go behind objects, and this might be something I will try, although I could avoid using it in my finished piece. The next problem was that when moving the camera, the light moves with it, when it should stay in place. I have some ideas of how to fix this, however. Lastly, I need a way to give the light a Z axis, so that it can appear to travel in 3D space, and move away from the camera as the people do, and shrink due to perspective. I have not yet looked into this.

General updates

I have filmed a small piece of film for effects testing, and have been experimenting for a short while, and have found everything far easier to do than expected. When my new computer which has more stuff than this one arrives, I should be able to get a move on. I'll upload my demo when it is done, and plan further from there.

Several demos have now been completed, and I have gained the ability to understand the basics of After Effects

I have chosen a parkour actor

The story has been finalised

I'll replace this as I go.

Helsinki

This is the chosen song, and I will start by studying the structure of the song.

Intro

0.00 - Intro
0.15 - bass enters

Verse 1

0.30 - Singing enters
0.31 - Drums enter

Pre - chorus 1

1.01 - bass, synth and vocal changes, tension built

Chorus 1

1.17 - guitar riff, tension released

Verse 2

1.32

Pre - Chorus 2

2.10

Chorus 2

2.26

Verse 3 (different style to 1 and 2)

2.42

Pre - Chorus 3

3.28

Chorus 3

3.58

Outro

4.12 - same as Chorus 3, but with vocals
4.29 - everything but synth cuts out
4.37 - song ends

I need to cut the length of the song down to under 4 minutes, aiming for something around 3ish for the time being (although I will hopefully make a full version too).

Things to possibly cut:

1. A bar from the intro , taking off about 7 seconds
2. Cut out a section of vocals at the 4.12 point
3. An entire verse - possibly the second, as well as taking out the linking chorus
4. Cut out about 5 seconds from the synth right at the end of the song
5. Remove chorus 2/3

Update: The song has been edited, and was first uploaded with the drafts of the video

Temple Scene

I have decided to use Temple Scene for my music video, for several reasons. Firstly, they have a really interesting sound which is perfect for the parkour/light mix. They have looked at light graffiti before too, their cover art for one of their albums was created using it. The band shares similar ideas to me, which I feel is useful for creating a video that both of us will be happy with.

Here is a record of the conversations between myself and the band, dating back from the 8th of July, explaining the choice of some things, such as choice of song and refining of ideas.

__________________________________________

- I asked about using one of their songs for a video -

__________________________________________

Hi Tom,


No problem doing a music video - as long as there are no restrictions on us using it if we like it! :-)


It would be great if you could use something from our new album (which you can hear at templescene.bandcamp.com - let me know if you want a code to download it). What kind of video are you thinking of doing?


Ric :)
__________________________________________

Brilliant, thanks a lot - I'd be honoured if you used the final product, although I can't guarantee it'll look professional, given the equipment we have to use and my experience, but I'll try my best. So far I'm just getting some initial ideas together, I've been looking into light graffiti, fire staffs and parkour (possibly merging the ideas). You can see my first ideas on my media studies blog, http://tkail2.blogspot.com/, where i post my ideas and everything media related. Look for the post "LIGHT AND FIRE EFFECTS" - there are some videos there that have really inspired me, and I feel that the light graffiti particularly would fit with your style of music, possibly combined with parkour - here are some videos from the group I would use - http://www.youtube.com/watch?v=-OUvNZidl-Y&feature=player_profilepage http://www.youtube.com/watch?v=IePjgBMWHbk&feature=player_profilepage. Search for "Norwich Parkour" to find more. If you have any ideas or issues, please let me know - any help at all is greatly appreciated - you understand your music much more than me. I would also need to cut down some of the song, we need to create videos of 2 - 4 minutes of length, and from my research, videos tend to get dull after the 3rd, so i might need to cut out a verse depending on the song. A code for the album would be really useful, if you could supply one - thanks, I'm currently listening to it. Thanks again.

Tom

P.S How come you aren't signed? The music sounds brilliant, and you've reached the top of pretty much every unsigned band resource!

__________________________________________

Hi Tom,


That sounds cool! Funnily enough, we've made the light graffiti connection too. The picture for the new album cover is a similar thing using sparklers. We found it on the net and the guy let us use it - we then inverted the colours, which is why it's blue-on-white. Also, in an effort to have something on YouTube, we edited one of our songs onto a BMW advert that uses similar effects (http://www.youtube.com/watch?v=KLXL1OGmt0I).


As for not being signed.... well I think it's a combination of not really going for record companies (we want to write songs for a living but we're not sure record companies are the way to go any more) and the fact that we're not a touring band (it's just the two of us in a home studio!). So basically we're just doing whatever we can do to promote our music online for £0 :-)


If you did use an edited version of a song (which we could do for you), would you also be able to make a full-length version, or would that add a huge amount of work?


Let me know if there are any songs you want to know more about (e.g. lyric meanings etc) or ideas you have or anything else you want to ask!


Ric


PS, go to templescene.bandcamp.com/yum and use the code - tough luck, go buy it! - to download the album :-)

__________________________________________

Ah, that's brilliant, I was going to ask about that, its a brilliant photo. I saw the BMW video while researching you, its really nice - was it not the official one?
I could try to make a full length version, but it really depends on what the video is like, there's no point throwing in random clips if it ruins the effect. I've spent the past few hours discussing some ideas with a friend who is far better than me at parkour (also directed the "Down" video I linked), and knows about how to use it to keep someone interested while watching parkour videos. We were thinking of attaching lights to our hands, and then using special effects to create a trail of light from it, or expanding on that idea in the direction of someone following a trail of light - this is from my idea of parkour leaving a mark on where it goes, which I think would work well. One related idea was that a man, sitting around was hit with this trail of light, which goes through him and outside, that he follows, doing parkour on the way. We have no ending for this yet although I was thinking the man could turn slowly into the light, starting with glowing and eventually becoming the trail. At the end, the trail would find the same man, creating a loop. This is a bit pretentious and hard to do though. Doesn't really make much sense either, although it would create a form of narrative, which keeps the video interesting, which is the biggest problem with parkour videos. Let me know what you think of that idea, and tell me any other ideas you might have about it! We also decided that fire would be too dangerous to create an interesting video with, and that stop motion light effects would take literally years to produce, which rules them out.
Cheers for the code, I'm now downloading the album - I'll ask more about specific songs later.
Tom

__________________________________________

Hi Tom


The original BMW ad had a different piece of music. I think it was a DJ Shadow re-interpretation of a Beethoven piece. We just put and edit of our music on there, and slightly re-edited the advert in iMovie so the timing works.


I like the idea of the video having some sort of narrative structure (even if it doesn't make obvious sense!). Otherwise it would be in danger of turning into one of those 'cool skater' videos except with Parkour :)


I'll ask Philippe for his thoughts too.


Ric


__________________________________________

Hey Ric

I've just finished listening to the album repeatedly, I love it. Helsinki is my favourite song on the album, and unless I'm wrong, the lyrics start with "I'll chase the light; from anywhere to anyplace" - this seems to fit in pretty much perfectly with the "following the thread of light" idea - could you send me the full lyrics please? (its also really nice to sing to :P) So far, I'm pretty convinced on doing a video based on that one, unless you have any objections? Could you also tell me the ideas behind the song?

Tom

__________________________________________

Hi Tom,


I was thinking the same thing, but I didn't want to influence you :D


Helsinki's probably the most visceral song on the album lyrically - i.e. it's not telling a specific story so much as trying to evoke an emotion. It's got a strong sense of running, of trying to break through to something new and reinvent yourself. So I think it's perfectly suited to the idea. Also, it's divided into pretty distinct sections sonically (the verse, the dropdown bridge, the instrumental chorus, the post-chorus drops) which can probably be mirrored some way visually to help the 'narrative', if you can call it that.


In terms of getting as much of the song as possible in, it occurred to me that the first 30 seconds (until the vocal/beat starts) doesn't need any real action. It can more-or-less be a black screen with occasional flashes of revealing light, not a million miles away from a thunder storm or the gun sequences in "equilibrium"... so for example where there's a rhythmic cue (like the "rat-tat-tat" on the high hat at the beginning before the main drums) the lights could come up and show something (sheet-lightning style) and then back to black... and maybe it could go back to this in the post-chorus drops... I don't know if I'm explaining very well, feel free to ignore this idea completely :-)

Anyway, here are the lyrics:


I'll chase a light from anywhere to any place
Waiting for all traces of the memory to fade away
I should never have given up a single part
Of who I was or who I am


I want to be where you are
I want to be where you are
Helsinki


I'll take my time
Take back my words and promises and start again
And I will look upon this day as my release
As my escape, as my one and only way to show
Who I was and who I am


I want to be where you are
I want to be where you are
Helsinki


We're at a distance I can't work out how to bridge
And we've gone too far to ever get it back again
Who cares about the others
Who cares what the others do
Who cares about the others
Who cares what the others do


I want to be where you are
I want to be where you are
Helsinki
I want to be everywhere
I want to feel everything
I want to be everywhere


__________________________________________

Hi Tom,


Just checking in to see if you've had any mord thoughts/progress with the video :-)

Ric

__________________________________________

Thanks for contacting - I'm currently testing the idea, I've taken a short amount of footage, and am now trying to create the light effect - I've found a good plugin for After Effects, and just need to try to use it for now - i'll send screens and a video when its done! I'm fairly sure I'll use Helsinki, although I'll need to trim it down, I'll let you know which bits when I've decided. I will try to keep my mind open for ways to extend the video for the full song too, ideally it would last the whole thing. After discussing with some people, we came up the with the idea of using several people, each with a different coloured light in the video, although I still haven't got a clear narrative. I found this video http://www.youtube.com/watch?v=9tQEQQo8Q9o which I thought had some really good ideas, such as the interesting colour scheme. I've also set a date for the project to end - hopefully it'll be finished by the start of September, but this is more a target than an expectation. If you've had any thoughts, please let me know - I'll try to keep you updated regularly!
Tom


__________________________________________

Hi Tom!


I was just having a read of your blog and I have to say, quite apart from my vested interest, it's actually really interesting! Very well written. From what you were saying about the lessons you have drawn from the various videos, I was wondering if you have ever seen the film Run Lola Run, which (although it does not involve Parkour) pulls together a lot of the themes you are writing about, and also throws in some interesting tricks, from the shocking red hair of the protagonist to the sudden brief changes in style and pace of filming that ramp up the tension.

Just thought you might find it interesting if you have never seen it! You can probably see chunks of it on Youtube although it's well worth seeing the whole thing.


Ric :)

__________________________________________

Hey!
Thanks for the praise of the blog - feel free to leave comments if there's anything you want to say about a post! I just saw the intro on youtube, it looks good, i'll rent it sometime. Anyway, I've was looking for videos with the same idea as us on youtube, and I found this, which is just perfect.
http://www.youtube.com/watch?v=ox78WEmXQB8&feature=related
They have some fantastic editing and effects, and are using the same plugin I will use, so hopefully mine will compare, although they have much better lighting and mise-en-scene than I can hope for. Either way, it's fantastic, I especially like it when they use two torches, and the while pool ball. One thing I noticed was that the torch really needs to be in the center of the body or somewhere pivotal to the trick, or it just looks wrong. That's something to think about anyway.

Better light in motion

This is what I want.

As with the other similar videos, the same plugin and effects are used, but this pretty much uses the same ideas as me too, swapping freerunning for snowboarding. I showed this to a few people, and am happy to report that the gimmick of light and extreme sports held their attention for the entire film, sometimes wanting to see it again. I think that these guys might have one-uped my idea, given their superior equipment and mise-en-scene, but I think that if I have tricks that are really impressive, my video could be as good as this, and I would be happy with that. I'm going to set this video as my target to beat.

As with the other more inspiring videos studied, this video uses slow motion, interesting shots (note the perfect backgrounds used) and the same shot distances as "Love Don't Let Me Go", but this video is much more a set of clips, although organised.

Light in motion

This is the first use of the idea I had in motion. After speaking to the creator of the video, I have found out (as I had from my own research) that a plugin called Trapcode's Particular is needed. I can find no other way to do this, so I will download the demo version and play with it on a demo piece of footage. The problems with the plugin are that it works best when created frame by frame, moving the camera makes it much harder and moving the light behind an object is hard to do as the light is just a layer on top of the video. On an unrelated note, I really like the use of colour, and when I get a reply from the creator, I should know how they did it. Also note the nice mise-en-scene. This is something I might have problems with.

Parkour music videos

If I chose to, I could use the Norwich Parkour Alliance in my video. This would be a brilliant resource, and I really like the idea of combining freerunning with the fire or light effects mentioned previously. Having spoken to a friend (Lukman Hussin) who has created many parkour videos (http://www.youtube.com/user/QuadDamaaaaage), we decided that fire would be too hard and too dangerous to work with easily mixed with parkour, but using effects to create the light graffiti is possible, although time consuming. Although given I have a holiday to do this in, time might not be an issue.

I will be studying various parkour videos to look at the techniques used, and the different styles achieved. If I can find proof that it would be: 

1. Possible to make a successful video that is not just a montage of clips,
2. Possible to make a visually interesting film (as visually interesting is the pivot of the videos)
3. Possible to make with the technologies available to me
4. Possible to create completely different styles, using ranges of music for a different effect

1. A generic parkour video.

This is the basic format of most parkour videos - get footage of performer(s), often shot with a static camera, rarely with any particularly interesting shots, and add music, sometimes cutting on the beat for a flowing effect. Although this video is interesting for a while, I found it got dull long before the 5 minute point. After asking a sample of 10 people when they began to loose interested, the result was 2.07. Most of them stopped watching because the tricks began to repeat. They also said that the video was not very visually engaging besides the parkour, and that it lacked a professional finish. However, given that it is a parkour video, not a music video, it is fairly successful. Linking the shots together with a simple narrative might have made the video much more based around the music, and would have created (in my opinion) a much better video, although it would have been harder to do, and defeats the original purpose of showing off the parkour group, so in this respect they made the right choice. Although I do not like the music used, I do acknowledge the use of a song with a strong beat, which can be synchronised with the film easily to make it more engaging and interesting.

This is another generic video, but the people I talked to found it much less interesting (as did I). We put this down to the lack of action, the tricks which most of the people thought they could have done (visually uninteresting tricks), bad pace when with the music (the music was too fast, and the tricks were too slow), people did not like the music cutting out and starting again, or the failed trick attempts.

This is a third generic video, but held my audiences attention much longer than the others. We put this down to the more interesting tricks, speed of the tricks and cuts (fast pacing holds audiences attention well) and the more interesting shots. However, the audience did start to get bored of the repetition of the tricks, even the more interesting ones, and thought that the music didn't fit the mood, but was much better than nothing, and helped to keep interest a bit. There were occasional cityscape shots that everyone thought were very good, something to keep in mind, and the more interesting backgrounds were generally appreciated, as were the more dangerous tricks.

Although these videos were all slightly different, they all used many of the same techniques, and had similar feels to them. I will now look at the more different videos, which break these conventions, and are what I will learn more from. These videos have given me an idea of what to avoid (film not cut to the beat, uninteresting tricks, slow cuts, lyrics unrelated to themes, create a professional look, do not destroy concentration by stopping half way though).

This is a video that I greatly admire, and watched all the way through, as did everyone I showed it to. It has replay value, and is interesting, due to the unique idea of constant falling. Constant is a key word - it never pauses, giving the audience no reason to look away, although it does repeat tricks sometimes. The music fits well, although is not cut to the beat (the clear beat is still there though, which helps keep the video interesting. Although the "Down" gimmick is very interesting, and holds most peoples attention all the way through, the slow rotation of the video, a small and easy to accomplish thing, is an enormous help to keeping the attention of the audience, and leaves a lasting effect on them (most people said the room was spinning after watching). The small narrative (man shoots himself and dies) is also a real help to the video, leading it somewhere, instead of giving it the 'random shots' feeling of the other videos. However, even though the video is very successful, there are many things to be improved, such as the variation of shots and tricks. besides the parkour, most the shots have nothing else interesting in them, and the lack of variation detaches the audience from the film. Close ups would have been very engaging, although might not have worked with the theme, and would be hard to put in easily. However, shots taken from under what the person jumps over, or the person jumping towards the camera could have made the video much more interesting too, but would have defeated the twisting effect, as the effect only works because the shot types are so similar. The use of new ideas, as well as a narrative, however small, have here been proved vital, and gave the video a very different feel from the others, starting to escape the 'random shots' effect they created. A purpose to the tricks is vital.

Another much more interesting video, I will mainly be looking at the start of this video. The film is cut to the beat, showing seasonal images (this artist often works with seasons), and very pretty shots. The beautiful piano over the soft, natural image combine to create a very beautiful intro, and I feel that this is very successful. The human/cat eyes and the boy playing with the water make the connection between parkour and nature (which is the theme of the video). The titles screen enters as the music fades, and a drum rolls. The shots in this fraction of a second are perfect, blurring the camera to follow close up movement, and a shot of the shoes to show the parkour is starting. However, the actual parkour is not visually that impressive, even if the nature creates pretty shots. This is the main letdown of the video. The song ends half way through, and another starts, this time much heavier, and the video starts again, in a more industrial setting. As the transaction is done fast, it can hold attention, but that tends to be where I get bored, especially as with the natural pretty shots gone, I find the video gets less interesting. The introduction is amazing though. I will need to make sure the song I choose has a slow, soft intro.

This video is interesting to me because it has looked at filming parkour videos in a very different way. Where the other videos just show unconnected clips so to make each different, holding the viewers attention, this video uses the same mise-en-scene all the way through, using advanced filming techniques to hold the viewers interest. Even though the video, when tested on a sample of 15 people, was deemed "repetitive" and "dull" after a certain point, the sample agreed that the video was very entertaining for the first few minutes (at least, after it got started, everyone thought the intro was much too long). The techniques used makes the video appear more professional, and appears much more as a film than a random collection of shots. The artist is showing off both his parkour and cinematography skills. The slow build up of the song up to the climax in the first two minutes is mirrored by the video - starting slowly and calmly, but with a note of tension - shown in the mise-en-scene by the hiding of the man's face, a very common idea used in films (off the top of my head; how Sylar is portrayed in the first season of Heroes). As the guitar enters, the camera begins to get less static, moving around the figure, faster and faster, as fast cut editing complements the effect. A variety of shot distances are used, a contrast growing more and more as the video becomes more interesting. The song draws to a climax, and a bus drives in front of the camera, and a song begins after the pause. This intro uses some brilliant techniques and has some great ideas that would work with a song with a long, powerful intro.
How the parkour is shot is the next point to analyse. An establishing shot is used to show the viewer the small space the man will use, and a blurry effect is used to hold interest in what would have been a dull shot. As before, the man is presented as powerful by the shot types, mainly low angle shots. The parkour begins, and we are thrown into the action (creating a contrast with the minimal action before) with engaging close ups, and shots being cut on the beat of the man swinging into us (the camera). The video uses fast cuts, dynamic camera movement, fast, impressive tricking, and interesting camera angles to engage the audience in a way familiar to them - the style reminds me a lot of how fight scenes are shot in modern films. However, although this technique is highly effective, it does not show off the parkour as naturally as some other videos, and the shaky camera is very annoying and disturbs the mood. It is worth mentioning the time effects used in the video - some shots are obviously sped up or slowed down for effect, which is worth me looking into.

As well as looking at these videos, I though it would be important to look at the videos for the (slightly disappointing) recent video game "Mirrors Edge", as it was produced by professionals, and so can give a different, and more polished angle of the genre.

Although this video focuses more on showing off the game, there are still many aspects of this video which are not seen in most parkour videos and are very interesting. The first of these is the first person effect. This video pulls it off perfectly, and engages the audience by making them think they are the character. Talking to people who have played the game, many of them complained of feeling ill after playing the game because of the motion, and after researching this further, discovered that the game's creators were forced to place a reticle in the center of the screen to prevent disorientation. The problem with this effect is that it is hard to achieve without shaking the camera, which can ruin the video, and this also places the camera at risk, which I would rather not do. However, it may be worth implementing for small amounts of the film. The game also uses a clear colour scheme, which is visually appealing, and can be used to contrast the character, or as the game does, highlight certain points. However, this would also be very hard for me to achieve, given that Norwich isn't very bright and colourful. The trailer uses a shot right at the start where a foot hits the ground. This is a very powerful effect, and also very interesting, and so I will consider using it for bringing in new characters (it hides their faces, making them ambiguous and interesting) or exaggerating jumps. There are also birds eye shots used, which are used to show off heights, which could be used to make my character more danger-defying, which is again interesting. The last effect I am interested in from the video is the use of slow motion, which is used to exaggerate and beautify shots, making them seen much more epic, and are often used to show off backdrops. In this case they are used for tension - as the character leaps off a building, unknown to us whether she will survive or not, the colours wash, the shapes blur, and the scene slows down, before cutting out. This slow-mo effect could also be very well edited in with the music for a much greater effect, as they have done slightly here.

This is the only 'proper' music video which uses a form of parkour that I could find, and is most interesting to me for this, and so I will need to study the camera work, editing and use of parkour very deeply. Although the video does not technically use much parkour, instead using tricking and freerunning most of the time, combined with a form of capoeira; I will probably not stick strictly to parkour, and mostly use freerunning, which is a mix of parkour and tricking, due to it being the most visually interesting while sticking with my theme of running.

The film is edited very quickly, often cutting twice a second, which adds to the dynamic of the sport, mainly using mid-shots and long-shots throughout, with the long-shots to set the scene and the trick, with mid-shots cut in to get closer to the action, now we are clear with what it is. Unlike the Oleg Vorslav video, the tricks still shown clearly, so that the audience can really see what is happening. As in several other videos, this one uses slow motion to exaggerate tricks, and also in this case to give the audience time to see several people performing at once, which is interesting, and something I have not seen so far. Note that the slow motion is used (as well as the biggest tricks) in the chorus, which is the most powerful part of the song. One thing I don't much like about the video is the washed out colour, which I think takes away a major part of the video. It was likely dulled down because it was not very colourful to begin with, and clashed with other, more colourful parts. The lens flares are very effective at replacing colour though.

The camera work is also interesting, although much more generic than I expected to find, with some shots very similar to those in the videos in the top of the post, although the main difference is the use of angles. Shots are taken from various angles throughout on a horizontal level, for the purpose of creating interesting shots, often with crab camera movement around the focal point. On a vertical level, most the shots used are at the same height as the subject, although sometimes a low angle shot is used to make the character seem more powerful or impressive, such as the backflip at 1.10, or the woman dancing at 1.18. One shot I really like is at 1.47, when the camera moves towards the group of people as they run towards the camera, because of how engaging and powerful this is, especially as they do it at the start of the chorus, which sets it up well. Before the chorus, there is a transition as the camera follows the pole, which clearly indicates a change, which is something else I will need to bear in mind. 

The tricks used are a mix between dance and freerunning, although very little running is done, possibly because it is not very interesting to watch without frequent tricks. The video starts off using exaggerated forms of dancing, which I think look a bit cheesy, although they do work and are interesting to watch. The tricks are all flowing, and often link with other peoples tricks to form a sort of duet. Flips and spins are used in abundance throughout, mainly stationary, and this is probably because of how interesting and entertaining these are to watch, because of how most people cannot do them. The more technical tricks, such as precision jumps, which can be seen in the first video, are not seen in here, because most people would not appreciate the skill, because they are not particually interesting to look at. A lot of break dancing is also used, possibly because of how easy it is to achieve on solid ground, but I will not use it as it breaks the flow of movement, which I think is vital to the pace of a video. Also note that more people are around when less interesting tricks are shown, as a form of compensation.

This list of videos has shown me that every point I highlighted at the start is achievable, and so I have decided to base my video around parkour. I also choose to do this as it is an area not yet exploited, although Mirrors Edge shows that people are starting to. I will look into how best to combine parkour with light graffiti and use the effects together, with the things I have learnt from studying the above videos.

The really big list of things to do:

Cut to the beat.

Movement in camera, although not vital, is important in helping the flow, and can make the video feel as dynamic as the parkour. Hard to create effects with though, so to be used in limitation.

Use a variety of tricks - try not to repeat.

Use an interesting and pretty mise-en-scene if possible - this is limited in a city, but the cityscape and the sky can look good.

A very simple narrative can tie a parkour film together well, and can lead the characters to places.

Tricks must look impressive - the audience will not appreciate the more technical tricks.

Pacing and partnership are important - make sure the style/mood/speed of the video fits that of the song.

Do not let the music stop at any point, it ruins the effect (unless done for a specific effect, and this is very hard to do well) and ruins the flow and distracts the viewer, giving them a place to stop watching (in a music video, the fast pace is how the audience can become hooked, and even a small gap can allow the viewer to stop. Same goes for changing song midway through the video, although this won't be a problem given my task.

Failed tricks must not be used, even if part of the plot, unless it can be well justified and the repercussions worked out.

Fast editing and pace are vital, music videos are created to appeal to the average short attention spanned person.

Try to avoid repeating tricks - a few repeats are fine, but avoid basing a video around one trick.

The more dangerous the tricks, the more appealing they are to the audience.

Try to (at least loosely) relate lyrics to the film.

Professional look is important. Do not use tacky or cheap effects, and film with good lighting using a good camera. Pay attention to framing.

Try to have something about the video which leaves the audience thinking about it afterwards.

Use a variation of shot types.

Use engaging camera angles and shot types to make the audience feel part of the action.

Each trick must have a purpose, not just random tricks for the fun of it.

The intro to the film does not need to be entirely relevant - it can be used to set the scene and move with the intro to the song, creating a contrast between the intro and the rest of the video.

The mood of the mise-en-scene must suit the song.

Intro must not be too long - 15 seconds is plenty, even for a 4 minute video.

Create tension in the verse, which is released in the chorus. This can be done with music and video.

Establishing shots should be considered.

Consider representation of the character.

Smooth camera movement is a must.

Playing with speed can be useful - some tricks are too fast to see properly at normal speed, as well as the various effects that can be achieved by changing the speed.

Colour schemes can give the video an edge, and a professional look.

Use shots to show off heights etc. These can be faked.

Lens flares can give a lot to the composition of a shot, but also take a lot away, as well as doing several other things to a video.

Use transitions in the video to show contrast between movements in the song.

More people can make simple tricks interesting.

Ensure tricks (as well as the shots and editing) all flow together smoothly.

Possible songs to use

http://templescene.com/?p=306
Scroll down to "Ordinary People" and "Rewind" and "The Story"

Actually, most of their stuff. I've just sent an email to them asking if they really are unsigned, as their Myspace claims. They have long tracks, which is the only let down, as I expect to find that short music videos are the most effective, based on my own experiences of watching MTV or variants. I could try to cut them down though. Temple Scene have a very particular sound, which although I do not think would suit fire dancing, could work well with light graffiti, and possibly parkour (I've not looked at conventions of the videos yet)

I also like about half of "Harperactive"'s songs, which is annoying, because there are about 10 albums worth, and until I have chosen from my ideas, and tested them to make sure they are possible, choosing will be hard. I'll list them when I get home. Also "Dataphiles" have some good stuff, and have photos of light graffiti on their space http://blog.dataphilesmusic.com/, which would fit perfectly with my ideas for a video, and so relates. Again, these are long, but can be cut down easily enough. These two bands are much more electronic than Temple Scene, creating a very different effect, although I will need to listen further to decide whether they would suit the ideas I have (the robot suit idea would work though).

Initial video ideas

My first idea, influenced mainly by the video for Jason Forrest's War Photographer, would be to create a video that uses over-the-top awesomeness and originality to make video that engages the audience though comedy and its uniqueness. Influenced greatly by my love for Daft Punk, ideas for this include:

• People in cardboard robot suits dancing the robot
• People in cardboard robot suits fighting in front of fire
• People in cardboard robot suits rockin' out
• The evolution of the People in cardboard robot suits (starting with making fire from sticks up to using humans for processing needs?)

I like the idea of using old CRT monitors and TVs for the heads.

I'm also considering making a more serious video with the same ideas, much like Daft Punk's "Da Funk" video.

My next idea is to have someone performing a complex choreographed dance, in a dark room/ at night, while holding fire or lights, so the lights create intricate patterns. The best examples for this are these:

The second video uses light graffiti, which I would love to use, but is near impossible to film due to how it is done, meaning without using special effects, it would need to be done using stop motion, which would take a very long time, and a lot of tedious editing. Each of these videos gives off a different feel - the first video's fire dance would be best suited to a hardcore dance/trance/electronica song, due to the speed and intensity of the dance, whereas the second uses cheerful, light sound for it's music, which is perfectly suited to it, although this is mainly due to the themes. If I were to try light graffiti, I would need to explore the effect of narrative, colours, mise-en-scene and editing on the overall mood the video gives off, and use a song to suit the effect.

Another idea is to use the flow time to create new and interesting effects, such as in "The Scientist" by Coldplay, filmed entirely in reverse, revealing the narrative backwards as well. Coldplay's video "Yellow" also plays with time. While the singer walks towards the camera (which he does for the entire length of it), the time of day visibly changes from morning, to day, and eventually to sunset.

Quick thoughts regarding music genre for my music video

Because it is much easier to make a start on idea planning with an idea of the genre to be used, I have decided to use electronic music, and all the sub-genres surrounding it. I have chosen these because of the diversity of the existing music videos, and the lack of lyrics in most, allowing for true creative freedom on my part, instead of being restrained by the story behind the lyrics. Electronic music also often has a clear and powerful beat, making it easy to edit to, emphasizing the power of it. There is also a lot of high quality electronic music on the Internet, making finding a good band easier, especially thanks to the remixing community, who constantly edit, sample and remix other songs, creating a wide diversity in the range of music available.

4. Nick Cave & Kylie Mingoue - Where The Wild Roses Grow

This video in many ways is very similar to Jamiroquai's video "Virtual Insanity", in that the beautiful seeming video tells a darker story, which reveals itself in small parts until the ending, which explains what has happened, making the watcher want to watch the video again with a different understanding. In this case, director Rocky Schenck, who as well as being a painter and photographer in his own time, often produces visually beautiful videos, and this is no exception, using a gold colour scheme and a special effect (diffuse glow) which exaggerates the whites of the video, creating a soft reflection in everything. The camera movement is very smooth and gentle, which is reflected the mood the video gives off. All this gives the video a soft, beautiful, dreamlike feeling, which is exactly what the director wants the video to feel like before looking deeper.

The video is about two characters, played by Kylie Minogue and Nick Cave, Elisa Day and an unnamed man respectively. Cave's character is stunned by Elisa's beauty, so visits her home a presumably share a night together. He brings her a red rose Day two, and asks her if she'd like to see where it comes from "If I show you the roses, will you follow". The next day, he takes Elisa to the river, kisses her, and kills her with a rock. He then places her in the river where the beautiful rose he picked grew, in the pose of Millais' painting Ophelia.
He places the rose in her mouth, and leaves.

The video gives away none of this, besides small hints. The ending is the only point when the reader can acknowledge what has just happened, and are then more likely to listen to the lyrics and understand the true story behind the video, as listening to lyrics while watching a music video for the first time is very difficult. The video gives us no idea to suspect that she lies dead in the river before the hints, which start too small for us to suspect anything. The first of these is when he picks up a rock, which seems to be the focus of the shot. This stood out for me when first watching the video, mainly due to how out of place it seemed, making much more sense after a second time. The next is a snake crawling over her (presumably dead by this point) body, which relates the video to The Garden of Eden, a beautiful garden which is ruined by humanity, and it is tainted in this too. The snake in the garden has the connotations of deceitfulness, which fits the video, which lies to us about the true nature of the narrative until the end. The next hint at this is a millipede, which crawls around her hair, carrying the same connotations as the snake. These both suggest to us that her body is now a part of nature, returning to the earth. It is only when he shuts her eyes at the end of the video that her death is made clear, the perfect image of the garden ruined, and the rock explained.

If I were to create a visually beautiful video, I would use techniques such as the slow camera movement and diffuse glow effect. I also like how the narrative reveals itself, only truly making sense after repeated viewings, and the order of the narrative, starting and ending with the story being told by Elisa in heaven. Playing with the time of events is a more interesting way of getting a simple story (which is normally all there is time for in a music video) told.